JP2019012323A - Vehicle controller - Google Patents

Abstract

To perform fusion properly even when there a gradient difference is generated between an object and an own vehicle.SOLUTION: An ECU 10 is applied to a vehicle comprising a radar device 21 and an imaging device 22 as body detection sensors detecting a body present at a periphery of an own vehicle 50. An ECU 10 comprises: a first acquisition part which acquires, based upon a detection result of the radar device 21, a first position of a body specified with a distance and an azimuth based upon the own vehicle 50; a second acquisition part which acquires, based upon a detection result of the imaging device 22, a second position of the body specified with the distance and azimuth based upon the own vehicle 50; a gradient calculation part which calculates a gradient difference as a difference in road gradient between the body and own vehicle 50; a correction part which corrects, based upon the gradient difference, the distance of the second position to the body; and an identification part which identifies, based upon the first position and the second position corrected by the correction part, whether the body detected by the radar device 21 and the body detected by the imaging device 22 are identical.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a vehicle control device applied to a vehicle including a radar device and an imaging device.

  The radar position based on the detection result of the radar device and the image position based on the detection result of the imaging device are collated, and if it is determined that they are due to the same object, the radar position and the image position are fused. A technique for generating a new object position (fusion position) is known.

  For example, in the target detection device of Patent Document 1, when the radar position and the image position are in a predetermined proximity relationship, it is determined that the radar position and the image position are due to the same object. Both the radar position and the image position are acquired as relative coordinate positions where the vehicle width direction of the host vehicle is the X axis and the traveling direction of the host vehicle is the Y axis when the host vehicle is the origin. In this case, the radar position and the image position are specified based on the distance from the own vehicle to the object and the direction of the object with reference to the own vehicle.

JP 2014-122873 A

  Incidentally, the distance from the host vehicle to the object at the image position is calculated based on the vertical position of the object in the captured image. However, the road gradient (inclination) cannot be considered from the captured image. Therefore, when a relative gradient difference occurs between the object and the host vehicle, an error may occur in the distance from the host vehicle to the object at the image position. As a result, there is a possibility that a deviation occurs between the image position and the radar position, and proper identification determination cannot be performed.

  The present invention has been made in view of the above problems, and a main object of the present invention is a vehicle control device capable of appropriately performing fusion even when a gradient difference occurs between an object and the host vehicle. Is to provide.

The first means is
The present invention is applied to a vehicle including a radar device (21) and an imaging device (22) as an object detection sensor that detects an object existing around the host vehicle (50).
A first acquisition unit configured to acquire a first position of the object specified by a distance and a direction based on the own vehicle based on a detection result of the radar device;
A second acquisition unit configured to acquire a second position of the object specified by a distance and an orientation based on the host vehicle based on a detection result of the imaging device;
A gradient calculating unit that calculates a gradient difference that is a difference in road gradient between the object and the host vehicle;
A correction unit that corrects the distance of the second position to the object based on the gradient difference;
Based on the first position and the second position corrected by the correction unit, it is determined whether or not the object detected by the radar device and the object detected by the imaging device are the same object. The same determination unit;
It is characterized by providing.

  Whether or not the object detected by the radar device and the object detected by the imaging device are the same object (identical determination) is determined based on the radar position and the image position acquired based on the detection result of each device. It is done using. However, since the road gradient is not taken into account in the object detection by the imaging device, for example, when a gradient difference occurs between the object and the host vehicle, an error may occur in the distance to the object at the image position. As a result, there is a possibility that appropriate identity determination cannot be performed.

  In this regard, in the above configuration, a gradient difference that is a difference in road gradient between the object and the host vehicle is calculated, and the distance to the object at the second position (image position) is corrected based on the calculated gradient difference. Then, the same determination is performed based on the first position (radar position) and the corrected second position. In this case, by correcting the distance to the object at the second position based on the gradient difference, the second position can be acquired with the road gradient taken into account. Then, by using the first position and the corrected second position, an appropriate identity determination can be performed. Thereby, even if it is a case where a gradient difference arises, a fusion target can be produced | generated appropriately.

The block diagram which shows schematic structure of a vehicle control apparatus. The figure for demonstrating a radar position and an image position. The figure which shows the object detection by the imaging device in an uphill gradient path. The figure for demonstrating correction | amendment of an image search area | region. The figure which shows the relationship between correction amount, gradient difference, and the up-and-down position of a captured image. The figure for demonstrating correction | amendment of an image search area | region. The figure which shows the relationship between correction amount, gradient difference, and the up-and-down position of a captured image. The flowchart which shows the process sequence of the same determination.

  FIG. 1 shows a pre-crash safety system (hereinafter referred to as PCSS: Pre-crash safety system) to which a vehicle control device is applied. PCSS is an example of a vehicle system mounted on a vehicle. When an object existing around the host vehicle is detected and the detected object may collide with the host vehicle, the host vehicle collides with the object. The avoidance operation or the collision mitigation operation is performed.

  A host vehicle 50 illustrated in FIG. 1 includes a radar device 21 and an imaging device 22 as object detection sensors, a navigation device 23, an ECU 10, and a driving support device 30. In the embodiment shown in FIG. 1, the ECU 10 functions as a vehicle control device.

  The radar device 21 detects an object in front of the host vehicle using a directional electromagnetic wave (exploration wave) such as a millimeter wave or a laser, and its optical axis is at the front of the host vehicle 50. It is attached so that it faces. The radar device 21 scans a region extending in a predetermined range toward the front of the host vehicle with a radar signal every predetermined time, and receives an electromagnetic wave reflected from the surface of the object in front of the host vehicle to receive the relative position ( Detection position), relative speed, etc. are acquired as object information. The acquired object information is input to the ECU 10 at predetermined intervals.

  The imaging device 22 is an in-vehicle camera, and is configured using, for example, a CCD camera, a CMOS image sensor, a near infrared camera, or the like. The imaging device 22 is attached to a predetermined height (for example, near the upper end of the windshield) in the center of the host vehicle 50 in the vehicle width direction, and images a region extending in a predetermined angle range toward the front of the host vehicle from an overhead viewpoint. The captured image is input to the ECU 10 at predetermined intervals. In the present embodiment, the imaging device 22 is a monocular camera.

  The navigation device 23 provides the ECU 10 with road information of the road on which the host vehicle 50 travels. For example, the navigation device 23 includes a memory for recording map information and a position specifying unit for specifying the position of the host vehicle 50 on the map by using positioning information transmitted from a GPS (Global Positioning System) satellite. Yes. And the navigation apparatus 23 refers to the road information around this vehicle position based on the vehicle position on the specified map. Then, the referred road information is transmitted to the ECU 10. The road information is, for example, road gradient information (such as an inclination angle), information indicating the presence or absence of an intersection, and the like.

  The driving support device 30 is an alarm device that emits an alarm sound to the driver, or a brake device that decelerates the vehicle speed of the host vehicle 50, and performs a collision avoiding operation or a collision reducing operation with an object according to a control command from the ECU 10. If the driving assistance device 30 is a brake device, the automatic brake is activated when the possibility of collision with an object increases. Further, if the driving support device 30 is an alarm device, an alarm sound is emitted when the possibility of collision with an object increases.

  ECU10 is comprised as a known microcomputer provided with CPU and various memory (ROM, RAM), and implements control in the own vehicle 50 with reference to the arithmetic program and control data in memory. The ECU 10 recognizes an object based on the detection result (object information) input from the radar device 21 and the detection result (captured image) input from the imaging device 22, and operates the driving support device on the recognized object. PCS with 30 as a control target is implemented. Below, the PCS control which ECU10 implements is demonstrated.

  The ECU 10 acquires the radar position of the object specified by the distance and azimuth relative to the host vehicle 50 based on the object information. In addition, the image position of the object specified by the distance and direction with reference to the host vehicle 50 is acquired based on the captured image. Then, based on the radar position and the image position, the same determination is made as to whether or not these positions are due to the same object.

  Here, the radar position and the image position will be described with reference to FIG. Each position is acquired as a relative coordinate position where the vehicle width direction of the host vehicle 50 is the X axis and the traveling direction of the host vehicle 50 is the Y axis when the host vehicle 50 is the origin.

  The ECU 10 first acquires a radar detection point Pr as an object detection position by the radar device 21. The radar detection point Pr includes a relative distance r1 from the host vehicle 50 to the object and an azimuth θr of the object with respect to the host vehicle 50. Further, the image detection point Pi is acquired as the detection position of the object from the captured image. The image detection point Pi includes a relative distance r2 from the host vehicle 50 to the object, and an azimuth θi of the object with reference to the host vehicle 50. The relative distance r2 is calculated based on the lower end of the object detected in the captured image.

  Subsequently, the ECU 10 acquires a radar search region Rr based on the radar detection point Pr. The radar search region Rr is a region having a width corresponding to an assumed error that is set in advance based on the characteristics of the radar device 21 in each of the distance direction and the azimuth direction with reference to the radar detection point Pr. For example, the radar search area Rr is set as an area that includes the radar detection point Pr (r1, θr) and extends in a predetermined distance range and a predetermined azimuth range.

  Further, the ECU 10 acquires the image search area Ri based on the image detection point Pi. The image search region Ri is a region having a width corresponding to an assumed error set in advance based on the characteristics of the imaging device 22 in each of the distance direction and the azimuth direction with the image detection point Pi as a reference. For example, the image search area Ri is set as an area that includes the image detection point Pi (r2, θi) and extends in a predetermined distance range and a predetermined azimuth range. In the present embodiment, the radar search area Rr is used as the radar position, and the image search area Ri is used as the image position.

  Then, the ECU 10 determines whether the object detected by the radar device 21 and the object detected by the imaging device 22 are the same object based on a predetermined determination condition. Here, as a determination condition, it is determined whether or not there is an overlapping region between the radar search region Rr and the image search region Ri. For example, in FIG. 2, since there is an overlapping area between the radar search area Rr and the image search area Ri, these objects are determined to be the same object. In such a case, the radar position and the image position are merged to generate a fusion position that becomes a new position with respect to the object. In FIG. 2, the accurate relative distance r1 of the radar detection points Pr (r1, θr) and the highly accurate azimuth θi of the image detection points Pi (r2, θi) are fused to form a fusion position. A fusion detection point Pf (r1, θi) is generated.

  Then, the ECU 10 determines whether or not the host vehicle 50 may collide with an object that has undergone the same determination. Specifically, it is determined whether or not the lateral position of each object belongs to the collision prediction area that is subject to collision avoidance control. In such a configuration, when the lateral position of the object belongs within the collision prediction area, it is determined that there is a possibility that the object and the host vehicle 50 collide.

  When it is determined that there is a possibility of collision between the object and the host vehicle 50, the ECU 10 calculates a collision margin time TTC (Time to Collision) for the object. Then, the driving support device 30 is operated based on the calculated TTC and the operation timing of the driving support device 30. Specifically, when the driving support device 30 is a brake device, if the TTC is equal to or lower than the operation timing, control is performed to reduce the collision speed by operating the automatic brake. Further, when the driving support device 30 is an alarm device, if the TTC is less than or equal to the operation timing thereof, an alarm is given to the driver by operating a speaker or the like. By such PCS control, avoidance or mitigation of the collision between the object and the host vehicle 50 is achieved.

  Incidentally, the distance from the host vehicle 50 to the object at the image detection point Pi (relative distance r2 in FIG. 2) is calculated based on the vertical position of the object in the captured image. Here, FIG. 3 shows a situation in which the host vehicle 50 travels on a flat road (inclination angle zero), and the preceding vehicle 60 as an object travels on an upward slope road having an inclination angle A. Under such circumstances, since the position of the preceding vehicle 60 is positioned above the position of the host vehicle 50, the preceding vehicle 60 is detected at a higher position in the captured image. However, since the road gradient cannot be taken into consideration from the captured image, the position of the preceding vehicle 60 is detected as the position 70 according to the image detection point Pi based on the captured image. Therefore, an error (ΔD) occurs between the distance D2 based on the image detection point Pi and the distance D1 based on the actual position of the preceding vehicle 60. In this case, the distance D2 is larger than the distance D1. (D2> D1). On the other hand, even in a situation where the preceding vehicle 60 travels on a downhill road unlike FIG. 3, an error (ΔD) between the distance D2 based on the image detection point Pi and the distance D1 based on the actual position of the preceding vehicle 60. Will occur. In this case, the distance D1 is larger than the distance D2 (D2 <D1). When the distance error (ΔD) occurs as described above, the position of the image search area Ri set based on the image detection point Pi is shifted in the distance direction, and the same determination may not be performed properly.

  Therefore, in the present embodiment, a relative gradient difference Δθ between the position of the object and the position of the host vehicle 50 is calculated, and the position of the image search region Ri in the distance direction is corrected based on the calculated gradient difference Δθ. Then, the same determination is performed based on the radar search region Rr and the corrected image search region Ri. That is, by using the image search area Ri that takes into account the road gradient, even if the gradient difference Δθ occurs, the same determination can be appropriately performed.

  The ECU 10 calculates a gradient difference Δθ that is a difference in road gradient between the preceding vehicle 60 and the host vehicle 50. In this case, the ECU 10 acquires gradient information of the host vehicle 50. Specifically, the inclination angle θ1 of the host vehicle 50 is acquired based on the current location of the host vehicle 50 acquired by the navigation device 23 and the road information at that point. In the configuration in which the host vehicle 50 includes a gyro sensor or an inclination angle sensor, the gradient information of the host vehicle 50 may be acquired based on the detection values of the sensors.

  Further, the ECU 10 acquires gradient information of the preceding vehicle 60. The gradient information of the preceding vehicle 60 is acquired based on road information at a position near the current location of the preceding vehicle 60. The position near the current location of the preceding vehicle 60 is calculated based on, for example, the distance to the preceding vehicle 60 at the image detection point Pi and the current location of the host vehicle 50. However, as shown in FIG. 3, when there is an uphill or downhill road ahead of the host vehicle 50, an error may occur in the distance of the image detection point Pi. The gradient information of the road in the predetermined range E including the distance of the image detection point Pi is acquired. For example, the average value of the inclination angles within the predetermined range E in the azimuth direction of the image detection point Pi is acquired as the inclination angle θ2 of the preceding vehicle 60. The predetermined range E may be variably set according to the distance of the image detection point Pi.

  The inclination angle θ1 indicates a positive value when the host vehicle 50 is an ascending slope road, and indicates a negative value when the host vehicle 50 is a descending slope road. In addition, the inclination angle θ2 indicates a positive value when the preceding vehicle 60 is an ascending gradient road, and indicates a negative value when the preceding vehicle 60 is a descending gradient road. For example, the values of the inclination angles θ1 and θ2 increase as the uphill slope becomes steeper.

  When the ECU 10 acquires the inclination angle θ1 of the host vehicle 50 and the inclination angle θ2 of the preceding vehicle 60, the ECU 10 calculates the gradient difference Δθ. For example, the ECU 10 calculates the gradient difference Δθ by subtracting the inclination angle θ1 of the host vehicle 50 from the inclination angle θ2 of the preceding vehicle 60. The gradient difference Δθ is a positive value when the gradient of the position of the preceding vehicle 60 is an upward gradient with respect to the gradient of the position of the host vehicle 50 (for example, in the case of FIG. 3). Therefore, the value is negative when the gradient of the position of the preceding vehicle 60 is a downward gradient.

  Below, correction | amendment of the image search area | region Ri by ECU10 is demonstrated. First, in FIG. 4, correction when the gradient of the position of the preceding vehicle 60 is an upward gradient with respect to the gradient of the position of the host vehicle 50 (Δθ> 0) will be described. For example, FIG. 4 shows a radar search region Rr and an image search region Ri acquired under the situation of FIG. In such a case, the ECU 10 corrects the image search region Ri based on the gradient difference Δθ. Specifically, first, the image detection point Pi is corrected so as to approach the host vehicle 50, and the corrected detection point CPi is obtained.

  The distance correction amount DA of the image detection point Pi is set based on the gradient difference Δθ (Δθ> 0) and the vertical position of the preceding vehicle 60 in the captured image. Here, it is considered that the deviation of the distance between the image detection points Pi increases as the gradient difference Δθ increases. Further, even with the same gradient difference Δθ, it is considered that the error in the distance of the image detection point Pi increases synergistically as the vertical position of the preceding vehicle 60 in the captured image becomes higher. Considering these points, in the present embodiment, the correction amount DA is set based on, for example, the correlation map shown in FIG. That is, in FIG. 5, the correction amount DA is calculated as a larger value as the gradient difference Δθ increases. Further, as the vertical position of the preceding vehicle 60 in the captured image becomes higher, the correction amount DA is calculated as a synergistically large value.

  Then, the ECU 10 sets a corrected image search region CRi based on the correction detection point CPi. Here, assuming that a normal image search region set based on the position of the correction detection point CPi is a region URi, the distance range Ca of the correction image search region CRi is set narrower than the distance range Ua of the region URi. For example, the distance range Ca is set by multiplying the distance range Ua by a predetermined ratio (for example, 60%) set in advance. On the other hand, the azimuth range of the corrected image search region CRi is the same as the azimuth range of the region URi. That is, the ECU 10 corrects the position of the image search area Ri toward the vehicle 50 based on the gradient difference Δθ, and corrects the distance range of the image search area Ri to be narrower than the normal distance range. On the other hand, the ECU 10 does not correct the radar search region Rr.

  On the other hand, FIG. 6 illustrates correction in the case where the gradient of the position of the preceding vehicle 60 is a downward gradient with respect to the gradient of the position of the host vehicle 50 (Δθ <0). In FIG. 6, for convenience of explanation, the same reference numerals are given to configurations similar to those in FIG. 4, and description thereof is omitted as appropriate. In this case, the ECU 10 corrects the image detection point Pi to the side away from the host vehicle 50, and acquires the corrected detection point CPi.

  The distance correction amount DA of the image detection point Pi is set based on the gradient difference Δθ (Δθ <0) and the vertical position of the preceding vehicle 60 in the captured image. In this case, the correction amount DA is set based on, for example, the correlation map shown in FIG. In other words, in FIG. 7, the correction amount DA is calculated as a larger value as the gradient difference Δθ becomes larger on the negative side. Further, the correction amount DA is calculated as a synergistically large value as the vertical position of the preceding vehicle 60 in the captured image becomes lower.

  Then, the ECU 10 sets a corrected image search region CRi based on the correction detection point CPi. In this case, the distance range Ca of the corrected image search region CRi is set narrower than the distance range Ua of the region URi. On the other hand, the azimuth range of the corrected image search region CRi is the same as the azimuth range of the region URi. That is, the ECU 10 corrects the position of the image search area Ri to the side away from the host vehicle 50 based on the gradient difference Δθ, and corrects the distance range of the image search area Ri to be narrower than the normal distance range. .

  Then, the ECU 10 determines whether or not there is an overlapping area between the radar search area Rr and the corrected image search area CRi. If there are overlapping areas, it is determined that they are the same object. On the other hand, when there is no overlapping area, it is determined that they are not the same object.

  The same determination process performed by ECU10 is demonstrated using the flowchart of FIG. This process is repeatedly performed by the ECU 10 at a predetermined cycle.

  In step S11, a radar detection point Pr is acquired as a detection position by the radar device 21. In step S12, the radar search region Rr is acquired based on the radar detection point Pr. In step S13, an image detection point Pi is acquired as a detection position by the imaging device 22. In step S14, an image search area Ri is acquired based on the image detection point Pi. Steps S11 and S12 correspond to a “first acquisition unit”, and steps S13 and S14 correspond to a “second acquisition unit”.

  In step S15, gradient information on the position of the host vehicle 50 is acquired. Specifically, the inclination angle θ <b> 1 is acquired by referring to road information based on the current location of the host vehicle 50 by the navigation device 23. In step S16, as the object, for example, gradient information of the position of the preceding vehicle 60 is acquired. Specifically, the inclination angle θ2 is acquired by referring to road information based on the vicinity of the current location of the preceding vehicle 60 by the navigation device 23. In step S17, the gradient difference Δθ is calculated based on the acquired θ1 and θ2. Here, the gradient difference Δθ is calculated by subtracting the inclination angle θ1 of the host vehicle 50 from the inclination angle θ2 of the preceding vehicle 60. If the gradient difference Δθ is a positive value, it indicates that the preceding vehicle 60 is an upward gradient with respect to the own vehicle 50, and if the gradient difference Δθ is a negative value, it indicates that the preceding vehicle 60 is a downward gradient with respect to the own vehicle 50. Show. Steps S15 to S17 correspond to a “gradient calculation unit”.

  In a succeeding step S18, it is determined whether or not the calculated Δθ is equal to or larger than the threshold Th1 and equal to or smaller than the threshold Th2. For example, the threshold value Th1 is set to zero or a negative value near zero, and the threshold value Th2 is set to zero or a positive value near zero. If step S18 is YES, the process proceeds to step S20 without correcting the image detection point Pi and the image search area Ri. In this case, for example, if Δθ is zero, there is no gradient difference between the preceding vehicle 60 and the host vehicle 50, and therefore it is considered that there is no distance error caused by the gradient difference in the image search region Ri.

  On the other hand, if step S18 is NO, that is, if Δθ is larger than the threshold Th2, or if Δθ is smaller than the threshold Th1, the process proceeds to step S19. In step S19, the image detection point Pi and the image search area Ri are corrected based on the gradient difference Δθ. Specifically, if Δθ is larger than the threshold Th2, as shown in FIG. 4 described above, the image detection point Pi is corrected to the side closer to the host vehicle 50, and the corrected image is based on the corrected correction detection point CPi. By setting the search area CRi, the image search area Ri is corrected. If Δθ is smaller than the threshold value Th1, as shown in FIG. 6 described above, the image detection point Pi is corrected to the side away from the host vehicle 50, and the corrected image search region CRi is corrected based on the corrected correction detection point CPi. Is set to correct the image search area Ri. Step S19 corresponds to a “correction unit”.

  In step S20, it is determined whether or not the same determination criterion is satisfied. Here, as a determination condition, it is determined whether there is an overlapping area between the radar search area Rr and the image search area Ri. If the image search area Ri is corrected in step S19, it is determined whether or not there is an overlapping area between the radar search area Rr and the corrected image search area CRi. In step S20, other conditions other than the overlapping region may be determined. For example, whether or not the absolute value of the difference between the radar TTC calculated based on the radar detection point Pr and the image TTC calculated based on the image detection point Pi or the corrected detection point CPi is less than a predetermined value B. You may judge.

  If step S20 is YES, the process proceeds to step S21, and the fusion detection point Pf is generated as the same object. On the other hand, if step S20 is NO, the process proceeds to step S22. In step S22, the radar detection point Pr and the image detection point Pi or the correction detection point CPi are maintained without generating the fusion detection point Pf. Note that step S20 corresponds to “same determination unit”. Thereafter, the ECU 10 determines the possibility of collision for each set detection point.

  According to the embodiment described in detail above, the following excellent effects can be obtained.

  In the above configuration, the gradient difference Δθ, which is the difference in road gradient between the preceding vehicle 60 and the host vehicle 50, is calculated, and the distance of the image detection point Pi and the distance direction of the image search region Ri are calculated based on the calculated gradient difference Δθ. The position was corrected. Then, the same determination is performed based on the radar search region Rr and the corrected image search region CRi. In this case, by correcting the position of the image search area Ri based on the gradient difference Δθ, the image search area Ri can be acquired with the road gradient taken into account, and thus appropriate identification determination can be performed. . Thereby, even if the gradient difference Δθ occurs, the fusion detection point Pf can be appropriately generated.

  Specifically, when the gradient of the position of the preceding vehicle 60 is an upward gradient with respect to the gradient of the position of the host vehicle 50 (Δθ> Th2), the position of the image search region Ri is corrected to be closer to the host vehicle 50. I tried to do it. Further, when the gradient of the position of the preceding vehicle 60 is a downward gradient with respect to the gradient of the position of the host vehicle 50 (Δθ <Th1), the position of the image search region Ri is corrected to the side away from the host vehicle 50. did. Therefore, it is possible to perform correction according to a situation in which an error in the distance of the image search area Ri occurs.

  When the gradient difference Δθ occurs between the object and the host vehicle 50, it is considered that the degree of error in the distance of the image detection point Pi differs depending on the vertical position of the object in the captured image. Considering this point, the correction amount DA is acquired based on the gradient difference Δθ and the vertical position of the preceding vehicle 60 in the captured image, and the distance of the image detection point Pi is corrected based on the correction amount. In addition to the gradient difference Δθ, an error in distance according to the vertical position of the preceding vehicle 60 in the captured image can be taken into account. Thereby, the distance of the image detection point Pi can be corrected appropriately, and consequently the image search area Ri can be corrected appropriately.

  When correcting the position of the image search area Ri to the side closer to the host vehicle 50 or the side away from the host vehicle 50, for example, if the distance range of the image search area Ri is widened, the same determination condition is easily established. Therefore, the accuracy of the same determination may be reduced. In consideration of this point, based on the gradient difference Δθ, the position of the image search region Ri is corrected to the side closer to the host vehicle 50 or to the side away from the host vehicle 50, and the distance range of the image search region Ri is reduced. Was corrected. In this case, by narrowing the distance range of the image search region Ri, the radar search region Rr and the corrected image search region CRi are less likely to overlap, and the same determination condition is less likely to be satisfied. Thereby, it can suppress that the conditions of the same determination are satisfied too much, and can suppress that the precision of the same determination falls by extension.

  Based on the distance from the own vehicle 50 to the preceding vehicle 60 and the map information, the gradient information of the preceding vehicle 60 and the own vehicle 50 is acquired. In this case, each gradient information can be acquired by referring to the map information at the point where the host vehicle 50 exists and the point where the preceding vehicle 60 exists. Thereby, the relative gradient difference Δθ between the position of the preceding vehicle 60 and the position of the host vehicle 50 can be calculated with high accuracy.

(Other embodiments)
In the above embodiment, the same determination is performed based on the radar search region Rr and the image search region Ri. Any configuration that performs the same determination based on this point, the radar detection point Pr, and the image detection point Pi may be used. For example, the radar detection point Pr and the image detection point Pi may be determined to be the same object when they exist within a predetermined distance. In this configuration, the ECU 10 corrects the image detection point Pi based on the gradient difference Δθ to obtain a corrected detection point CPi. Then, based on the fact that the corrected detection point CPi and the radar detection point Pr exist within a predetermined distance, the same determination is performed. The distance correction amount DA of the image detection point Pi is set based on, for example, the correlation map shown in FIG.

  In the above embodiment, when the inclination angle θ2 of the preceding vehicle 60 is acquired, the position near the current location of the preceding vehicle 60 is calculated based on the distance of the image detection point Pi. A position near the current location of the car 60 may be calculated.

  In the above embodiment, the image detection point Pi is corrected, and the image search area Ri is corrected based on the correction. That is, the distance of the image detection point Pi is corrected and the corrected image search region CRi is set based on the correction detection point CPi. However, the present invention is not limited to this. For example, the image search area Ri may be directly corrected without going through the correction of the image detection point Pi. In such a configuration, the ECU 10 sets the position correction amount DA in the distance direction of the image search region Ri based on the gradient difference Δθ and the vertical position of the object in the captured image. Then, the ECU 10 sets a corrected image search area CRi from the image search area Ri based on the correction amount DA.

  In the configuration for correcting the distance range of the image search region Ri, the predetermined ratio may be variable. For example, a predetermined ratio may be set according to the correction amount DA. In such a configuration, for example, the larger the correction amount DA, the smaller the predetermined ratio is set. In this case, by setting the distance range of the corrected image search region CRi to be small as the correction amount DA increases, it is possible to suppress the same determination condition from being satisfied excessively.

  In the above embodiment, the object is the preceding vehicle 60. However, the present invention is not limited to this, and for example, a bicycle or a pedestrian may be the target object.

  -In above-mentioned embodiment, it was set as the structure which implement | achieves appropriate collision avoidance control by applying ECU10 to PCSS and improving the precision of the same determination of an object. This may be changed and applied to, for example, an ACC (Adaptive Cruise Control) system that performs control for causing the ECU 10 to follow the preceding vehicle 60. In such a configuration, the same determination of the preceding vehicle 60 is performed with high accuracy, so that appropriate follow-up traveling control is realized.

  DESCRIPTION OF SYMBOLS 10 ... ECU, 21 ... Radar apparatus, 22 ... Imaging device, 50 ... Own vehicle.

Claims (7)

The present invention is applied to a vehicle including a radar device (21) and an imaging device (22) as an object detection sensor that detects an object existing around the host vehicle (50).
A first acquisition unit configured to acquire a first position of the object specified by a distance and a direction based on the own vehicle based on a detection result of the radar device;
A second acquisition unit configured to acquire a second position of the object specified by a distance and an orientation based on the host vehicle based on a detection result of the imaging device;
A gradient calculating unit that calculates a gradient difference that is a difference in road gradient between the object and the host vehicle;
A correction unit that corrects the distance of the second position to the object based on the gradient difference;
Based on the first position and the second position corrected by the correction unit, it is determined whether or not the object detected by the radar device and the object detected by the imaging device are the same object. The same determination unit;
A vehicle control device (10) comprising:   The correction unit acquires a correction amount based on the gradient difference and the vertical position of the object in a captured image captured by the imaging device, and based on the correction amount, the object at the second position is acquired. The vehicle control device according to claim 1, wherein the distance of the vehicle is corrected. The first acquisition unit acquires, as the first position, a first search region that includes a detection position by the radar device and extends in a predetermined distance range and a predetermined azimuth range,
The second acquisition unit acquires, as the second position, a second search region that includes a detection position by the imaging device and extends in a predetermined distance range and a predetermined azimuth range,
The correction unit corrects the position of the second search region based on the gradient difference,
2. The vehicle control device according to claim 1, wherein the same determination unit determines whether or not they are the same object based on the first search region and the second search region corrected by the correction unit. The correction unit corrects the position of the second search region to the side closer to the own vehicle when the road gradient of the position of the object is larger on the positive side than the road gradient of the position of the own vehicle. Correct the distance range of the second search area to be narrower,
The vehicle control device according to claim 3, wherein the same determination unit performs the same determination based on the presence of an overlapping area between the first search region and the second search region corrected by the correction unit. . The correction unit corrects the position of the second search region to the side away from the host vehicle when the road gradient of the position of the object is larger on the negative side than the road gradient of the host vehicle. Correct the distance range of the second search area to be narrower,
The vehicle according to claim 3 or 4, wherein the same determination unit performs the same determination based on the presence of an overlapping area between the first search region and the second search region corrected by the correction unit. Control device.   The correction unit acquires a correction amount based on the gradient difference and the vertical position of the object in a captured image captured by the imaging device, and determines the position of the second search region based on the correction amount. The vehicle control device according to claim 3, wherein correction is performed.   The gradient calculation unit acquires each gradient information of the object and the host vehicle based on a distance from the host vehicle to the object and map information, and calculates the gradient difference based on the each gradient information. The vehicle control device according to any one of 1 to 6.

Images